Agentic AI has moved beyond flashy demos and into everyday engineering work. In 2026, the real challenge is no longer whether developers and DevOps teams should use AI, but where agentic workflows actually help, where they introduce risk, and how to keep humans in control without slowing everything down.
In this demo-heavy session, Brian shows what “Agentic DevOps” looks like in real life across modern delivery platforms. You will see how GitHub Copilot’s coding agent can take on real engineering work from issue to pull request, with human review, testing, and approval still firmly in the loop. You will also see how GitHub Agentic Workflows extend that model beyond the IDE by automating recurring repository work such as issue triage, CI failure follow-up, documentation maintenance, and other background tasks that quietly drain team time.
But this session does not stop with GitHub. Brian will also show how the same agentic DevOps thinking applies to Azure DevOps environments, including practical patterns for using AI to accelerate backlog refinement, code changes, pull request review, pipeline troubleshooting, documentation updates, and operational workflows while still respecting governance, security, and release controls. Whether your team is all-in on GitHub, standardized on Azure DevOps, or actively working across both, the core challenge is the same: use AI where it creates measurable value, and put guardrails around it before it creates hidden debt.
You will leave with a clearer mental model for where coding agents fit, where repository-level agentic workflows belong, and how to build review, security, and measurement into AI-assisted delivery from the start, whether your team works in GitHub, Azure DevOps, or both.
You will learn:
- How Agentic DevOps patterns apply across GitHub and Azure DevOps workflows
- Practical guardrails for AI-assisted coding, automation, review, and release processes
- How to identify the metrics and warning signs that reveal whether AI is helping or creating hidden delivery debt